Math Meets Football: Is the New Extra Point a Game-Changer?

Punting on 3rd down would be overly conservative since you get 4 downs. I've always thought the NFL was a little to conservative. But then at the college level I am a fan of Georgia Southern. The Triple Option coaches(Paul Johnson at Ga Tech, Jeff Monken at Army...both formerly of Georgia Southern) their philisophy is if with that type of offense, if you can't move the ball 2-3 yards a play either with a Fullback dive, a QB keeper/midline option, or pitching it to an A-Back. Then we have bigger problems than just turning the ball over.

The NYT has put a bot together that looks at the numbers and determines whether you should go for it or not. They have a good article on it here: http://www.nytimes.com/2014/09/05/upshot/4th-down-when-to-go...

This definitely works at the highschool level, but becomes much more risky and nuanced in major collegiate BCS-level games, or especially in the NFL. On-sides kicks are incredibly difficult to recover when your opponent has been coached to be ready for them. We have a "hands team" for this specific reason.

Making things even more difficult, there is game film available to pretty much all college/NFL coaching staffs within hours of a game's completion. If you do stuff like this, your opponents are going to be ready. I know we (Clemson) are always ready for Georgia Tech and Paul Johnson to pull stuff like this, because that's what he does. They're able to make bank on this against lesser opponents, but it doesn't always fly against an equally matched (or better) foe.

You may have missed it in this article, but he does address variance (in the section explaining external incentives coaches include in their decision-making).

I believe you are rephrasing the oft-repeated adage, "The [stock] market can stay wrong longer than you can stay liquid."

As for your second point, a 2 point conversion is basically just like any goal line play which they practice many times and encounter in games often. So I would say they have maximized their ability for this play and it wouldn't change if they attempted it more.

The defense can recover a blocked Point after and return it for two as well. its been like that in the college game for as long as I can remember.

90+ yard plays are extraordinary, so the defensive score will not be a large factor.

The main thing is that they are tweaking the 1 point kicks to be slightly less certain, if there is an imbalance I think they will quickly adjust things again.

> - I see that the defense can now score two points by intercepting a two-point play and returning it. Doesn't this lower the utility of the two-point play as used in the article? Or have I misunderstood?

It's such an unlikely event that it shouldn't enter into anyone's calculation except in extreme cirumstances (e.g., you only need 1 point for certain victory, so why risk it). I doubt it will happen once per season.

> It isn't hard to imagine that teams will get better at executing two-point plays if they attempt more.

Teams will get better at defending it too. Teams might acquire personnel especially for the play, both on offense and defense.

> am I right in thinking that they are typically attempted by a team in a desperate position at the moment?

Often but not always desparate. Generally teams defaulted to extra points (1 point) and only used 2-point conversions when there was some certainty it would help. For example, if you are ahead by 1 point and there is time only for one more drive (offensive possession) by the other team, an extra point, which would increase your 1 point lead to 2 points, won't help you because the other team can score either 0 or >=3 points (i.e., scoring 1 or 2 points is impossible, except in a very unlikely event). You might as well go for 2 and earn a tie in case the opponent scores 3 (a field goal).

Also, a little vocabularly help so you look smart around the water cooler:

> one-point field goal

"extra point"

> match

"game"

> I see that the defense can now score two points by intercepting a two-point play and returning it. Doesn't this lower the utility of the two-point play as used in the article? Or have I misunderstood?

Probability of a 100-yard interception return is very low.

> Teams will get better at defending it too. Teams might acquire personnel especially for the play, both on offense and defense.

There really isn't a functional difference between a stopping a 2pt conversion and stopping a red zone drive from the 2 yard line. Defenses already know how to do this. It's a matter of play calling, reading, execution, and the raw will of each team at this point.

> Am I right in thinking that they are typically attempted by a team in a desperate position at the moment?

No, not at all. You can reach certain score combinations that favor a 2pt conversion. Say the score is 13-14, and the team with 14 just scored a touchdown. Now it is 13-20. That team might elect to attempt a 2pt conversion to raise their score to 13-22, which would require the other team to score TWICE in to get ahead or tie the game. If they don't make the 2pt, they're still up by a touchdown.

You also should account for the fact that coaches are often considering more than just points, and beyond that are often just irrational. Coaches should go for it on 4th-and-short more often than they do, but getting stopped on 4th down is demoralizing and makes the coach look foolish, which shifts their decision-making. Getting stopped on a 2-pt. conversion might also reflect poorly on the coach, so they are averse to calling that play even if it is statistically the right move.

I find this psychological side of coaching very interesting, and sports are a great venue for examining the psychology of decision-making. A great example of outcome-bias is the sports world's reaction to the Seahawks decision to throw the ball at the goal line in last year's Super Bowl. I would argue it was the proper decision as it maximized their chances of winning the game (it allowed them time for one passing play and two running plays, as opposed to just two running plays). Unfortunately for them, a very low percentage outcome occurred (interception), and that outcome has strongly biased how the play call has been judged.

I would argue, from the coach's perspective, the quantity to be maximized is not the probability of winning but the probability of retaining their job. 2-pt conversion calls are are judged harshly when they fail, however unfairly, and it's likely in a coach's best interest to just stay conservative and avoid scrutiny, even if that means playing sub-optimally.

I'd like to think the surprise two-point attempt from the 15 will be even more exciting now.

On the other hand, he is on the field for 55-60 snaps anyway. The additional 4 or so from the two-point conversions isn't going to change his risk profile much.

Good point, not to mention the fact that the defense can now return 2 point conversions for a TD. INT heavy QBs could be in trouble.

Point expectation is a good way to think of it, but consider the expected points of a 2pt (or 1pt) play vs the other 50 snaps a qb takes everywhere else on the field, where the expected points are less than 1.

An integer point difference on a single play is huge.

The risk to the QB is much smaller on a 2-point conversion. Often QBs are injured standing in the pocket, waiting for receivers to get open far downfield. On a 2-point conversion, that situation occurs much less: There are more running plays because of the short distance required (Seattle Seahawks fans might have somthing to say about that), and passing plays generally are over quickly because the receivers can't run more than 10 yards downfield.

Also, as others have said, 1 point is a high return on investment for one play.

On that note, if I were to make one change to the game, it would be to get rid of kick off and kick return. Yeah, it's exciting to watch, but as a player I hate it. It's the most danger you will ever put yourself in on the field. Concussions abound.

They already practice and develop plays for short yardage. The short down field at the goal line is different than first down conversions, but it isn't all that different than a red zone offense scheme.